Python 哈希算法實現圖片相似度比較

# coding: utf-8

import operator
from PIL import Image
import numpy as np
import cv2

"""圖片處理: 圖片截取、圖片相似度比對、哈希算法比對"""


def cmp_pic(pic1, pic2):
    """
    比對圖片相似度
    @param pic1:
    @param pic2:
    @return:
    """

    a = Image.open(pic1)
    b = Image.open(pic2)
    return operator.eq(a, b)


def image_interception(image):
    """
    圖片截取
    @param image:  目標圖片
    @return:
    """
    img = cv2.imread(image)
    print('圖片{}高度、寬度、通道數為：{}'.format(image, img.shape))  # (1792, 828, 3) 高度、寬度、通道數
    cropped = img[170:650, 0:900]  # 裁剪坐標為[y0:y1, x0:x1]
    cv2.imwrite(image, cropped)
    return image


def aHash(img):
    """
    均值哈希算法
    @param img:
    @return:
    """
    # 縮放為8*8
    img = cv2.resize(cv2.imread(img), (8, 8))
    # 轉換為灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # s為像素和初值為0，hash_str為hash值初值為''
    s = 0
    hash_str = ''
    # 遍歷累加求像素和
    for i in range(8):
        for j in range(8):
            s = s + gray[i, j]
    # 求平均灰度
    avg = s / 64
    # 灰度大於平均值為1相反為0生成圖片的hash值
    for i in range(8):
        for j in range(8):
            if gray[i, j] > avg:
                hash_str = hash_str + '1'
            else:
                hash_str = hash_str + '0'
    return hash_str


def dHash(img):
    """
    差值感知算法
    @param img:
    @return:
    """
    # 縮放8*8
    img = cv2.resize(cv2.imread(img), (9, 8))
    # 轉換灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    hash_str = ''
    # 每行前一個像素大於后一個像素為1，相反為0，生成哈希
    for i in range(8):
        for j in range(8):
            if gray[i, j] > gray[i, j + 1]:
                hash_str = hash_str + '1'
            else:
                hash_str = hash_str + '0'
    return hash_str


def pHash(img):
    """
    感知哈希算法(pHash)
    @param img:
    @return:
    """
    # 縮放32*32
    img = cv2.resize(cv2.imread(img), (32, 32))  # , interpolation=cv2.INTER_CUBIC

    # 轉換為灰度圖
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    # 將灰度圖轉為浮點型，再進行dct變換
    dct = cv2.dct(np.float32(gray))
    # opencv實現的掩碼操作
    dct_roi = dct[0:8, 0:8]

    hash = []
    avreage = np.mean(dct_roi)
    for i in range(dct_roi.shape[0]):
        for j in range(dct_roi.shape[1]):
            if dct_roi[i, j] > avreage:
                hash.append(1)
            else:
                hash.append(0)
    return hash


def classify_hist_with_split(image1, image2, size=(256, 256)):
    """
    通過得到RGB每個通道的直方圖來計算相似度
    @param image1:
    @param image2:
    @param size:
    @return:
    """
    # 將圖像resize后，分離為RGB三個通道，再計算每個通道的相似值
    image1 = cv2.resize(cv2.imread(image1), size)
    image2 = cv2.resize(cv2.imread(image2), size)
    sub_image1 = cv2.split(image1)
    sub_image2 = cv2.split(image2)
    sub_data = 0
    for im1, im2 in zip(sub_image1, sub_image2):
        sub_data += calculate(im1, im2)
    sub_data = sub_data / 3
    # print(sub_data)
    return sub_data


def calculate(image1, image2):
    """
    計算單通道的直方圖的相似值
    @param image1:
    @param image2:
    @return:
    """
    hist1 = cv2.calcHist([image1], [0], None, [256], [0.0, 255.0])
    hist2 = cv2.calcHist([image2], [0], None, [256], [0.0, 255.0])
    # 計算直方圖的重合度
    degree = 0
    for i in range(len(hist1)):
        if hist1[i] != hist2[i]:
            degree = degree + (1 - abs(hist1[i] - hist2[i]) / max(hist1[i], hist2[i]))
        else:
            degree = degree + 1
    degree = degree / len(hist1)
    return degree


def cmpHash(hash1, hash2):
    """
    Hash值對比
    @param hash1:
    @param hash2:
    @return:
    """
    n = 0
    # hash長度不同則返回-1代表傳參出錯
    if len(hash1) != len(hash2):
        return -1
    # 遍歷判斷
    for i in range(len(hash1)):
        # 不相等則n計數+1，n最終為相似度
        if hash1[i] != hash2[i]:
            n = n + 1
    return n

image_interception('11.png')
image_interception('11.png')

img1 = '1.png'
img2 = '2.png'

hash1 = aHash(img1)
hash2 = aHash(img2)
n = cmpHash(hash1, hash2)
print('均值哈希算法相似度：', n)

hash1 = dHash(img1)
hash2 = dHash(img2)
n = cmpHash(hash1, hash2)
print('差值哈希算法相似度：', n)

hash1 = pHash(img1)
hash2 = pHash(img2)
n = cmpHash(hash1, hash2)
print('感知哈希算法相似度：', n)

n = classify_hist_with_split(img1, img2)
print('三直方圖算法相似度：', n)